Performance Evaluation of Parallel Algorithm Using Multi-Core System

Sanjay Kumar Sharma

Department of Computer Science, Banasthali University, Rajasthan.


In this paper, we have designed the parallel algorithm to solve the numerical problems using multithreading on shared memory multiprocessors (or multi-core) system. I studied the several fundamental problems, including prefix sums, selection, and sorting, which often form the building blocks of other parallel algorithms. We provide the model for execution of multithreaded parallel algorithm. Solving the system of linear equations is one of the most useful numerical problems. System of linear equations appears in almost every branch of science and engineering. Many scientific and engineering problems can take the form of a system of linear equations. Cilk++ is a popular language and has support for parallel execution that is integrated into the language. The paper contains the sequential and parallel algorithm for solving the system of linear equations and comparison of results in both sequential and parallel.

Keywords: Parallel algorithm, Multithreading, Multi-core programming, Performance estimation.

     Back to TOC